While great gains have been made in post-processing methods for functional magnetic resonance imaging (fMRI), many studies in patient or pediatric populations continue to be limited by head motion artifact. Data from our group and others suggest that image distortions from magnetic susceptibility effects are head position dependent. Consequently, we hypothesize that acquisition and processing methods designed to reduce susceptibility-induced image distortions and signal voids and will improve the effectiveness of realignment movement correction methods. The long-term goal of this work is to develop image acquisition protocols, reconstruction approaches and post-processing methods that will improve on the effectiveness of head movement correction in fMRI. Our approach will be to first develop a comprehensive understanding of the impact on the effectiveness of motion correction of acquisition parameters, acquisition methods for minimization of susceptibility artifact, and susceptibility corrections in image reconstruction. This will be carried out mainly through extensive simulation studies of head motion and phantom studies using a computer-controlled motion phantom. The most promising approaches will be evaluated empirically in human fMRI studies. We will also develop acquisition and reconstruction methods that will explicitly allow the estimation and correction of motion-induced changes in the magnetic fields and the associated changes in image distortions. Finally, we will explore methods to model and account for residual error after image realignment. Methods that reduce magnetic susceptibility effects may show the largest improvement in motion correction in regions of the brain in close proximity to an air-tissue interfaces. These areas include the orbito-frontal cortex, inferior and medial temporal lobes, and the frontal pole - brain areas that are important for executive and emotional processing and have been implicated in affective disorders, substance abuse and numerous conditions. Our comprehensive approach - including acquisition, reconstruction and post-processing methods - will allow for nearly complete correction of large head motions, making fMRI a substantially more useful tool for studying patients with dementias and other psychiatric disorders and for studying pediatric populations.